Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 22:17:15 on February26,2002
Entry number 4402
Followups:
keyword=synch problems, decide to run unbuff
I got called in because with cosmics the busy signal from roc2 is getting stuck high.
We made some tests where I sent the readout time to an output register, it indicated
that sometimes the readout deadtime fluctuates to 400 microseconds, indicating
the buffers were filling up. Looking at the hit counts in the fastbus modules in
unbuffered mode shows a typical distribution like below (run 3273, for 50K events).
Slot Distribution of Extra Gates for ROC 2
0 0 0 77 83 85 90 86 93 92 98 97 92 98 107 114 113 100 114 112 115 8 8 8 8 0
Slot Distribution of Extra Gates for ROC 3
0 0 0 82 73 83 84 86 85 81 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Slot Distribution of Extra Gates for ROC 4
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
---------
Roc4 is clean. Rocs2 and 3 pick up a lot of extra noise in their 1877's (but not
uniformly) and Roc2's 1881M also has extra noise.
We decide to run in unbuffered mode. It is safe. We use the configuration "nobpm"
which has no roc14 in the readout and is unbuffered.
<< REMEMBER: use "nobpm" config >>